The Weather Company processes 20 terabytes of data each day.
How much is that? Two times the amount of printed material in the U.S. Library of Congress.
The company’s first goal with all that data is, not surprisingly, a weather forecast to keep people “safe and informed,” Domenic Venuto, general manager of the consumer division at The Weather Company, explained at the INMA/WWMN Big Data for Media Conference in New York City recently.
Venuto shared details about the company’s Weather Effects Index, an automatic targeting platform used to help advertisers target their messages to The Weather Channel’s audience. Weather insights are built into the company’s advertising targeting models, resulting in fascinating — and profitable — data.
“We take ... daily weather data over the last two years, and we correlate those with our clients’ sales data or foot data and we look for causation and correlation between an event and any number of weather variables.”
For example, warmer temperatures, not suprisingly, drive ice cream sales. But what really shoots sales up is the rate of change, not just the temperatures. Another interesting fact: Soda sales actually increase more in rainy and snowy winters than in other weather conditions.
These data points “give us insight into human behaviour that weather is driving,” Venuto says. “And then we put that information into our targeting engine so when we see those conditions arrive in the forecast ... we'll actually deliver messages based on that information because we want to amplify that human emotion, that weather-led action.”
His advice for news media companies working on their Big Data strategy:
- Don’t be afraid of data.
- Invest in the data science team.
- Remember, it takes time.
“Proceed with innovation at velocity.”